US20110026689A1 - Telephone call inbox - Google Patents
Telephone call inbox Download PDFInfo
- Publication number
- US20110026689A1 US20110026689A1 US12/726,932 US72693210A US2011026689A1 US 20110026689 A1 US20110026689 A1 US 20110026689A1 US 72693210 A US72693210 A US 72693210A US 2011026689 A1 US2011026689 A1 US 2011026689A1
- Authority
- US
- United States
- Prior art keywords
- call
- client
- consumer
- pattern
- caller
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/22—Arrangements for supervision, monitoring or testing
- H04M3/2218—Call detail recording
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/42221—Conversation recording systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/50—Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
- H04M3/51—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M2203/00—Aspects of automatic or semi-automatic exchanges
- H04M2203/55—Aspects of automatic or semi-automatic exchanges related to network data storage and management
- H04M2203/551—Call history
Definitions
- the present invention relates generally to telecommunications, and more particularly, to a method and system for creating an e-mail-like telephone call inbox of transcribed telephone calls between consumers and a business.
- the client When a business (hereinafter, the “client(s)”) uses pay per call advertising, the client may receive a number of calls from consumers seeking their products or services, etc. When business is good, the number of calls may be overwhelming and too time intensive to monitor. What is desirable to monitor includes determining the number of consumers and extracting significant data from consumer interactions, such as the number and type of products sold, the state or phase of a complicated transaction, or the type of consumer that is interested in a particular product or service.
- the client may subscribe to an audio recording and or transcription service.
- the client may receive, for example, 30 phone calls a day, and playing the recorded phone messages takes on average five to ten minutes apiece (or a similar amount of time in transcribed form) it may take three hours to listen/read all of the calls, which becomes tedious.
- the client may receive only the caller ID number and a short string identifying the name of the consumer on their telephone, which provides little information about the consumer.
- the client also needs to manually and mentally extract relevant information from each call, yet may have no tools for finding patterns in the calls that may affect their business.
- client may refer to any person or organization that may employ the method and/or system of the present invention, which may include, but is not limited to, an individual, a non-profit organization such as a university, and a for-profit business.
- the term “consumer” may refer to any person or organization that may call into the system using the method of the present invention, which may include, but is not limited to, an individual, a non-profit organization such as a university, and a for-profit business.
- the telephony server When a call is received from a potential consumer at a telephony server, the telephony server completes the call to a client, bridges the call, and begins recording the call. The identity of at least one consumer calling a client is extracted from the call. As part of the identification process, non-consumer fraudulent call data is automatically filtered out of the call (stream).
- a call is transcribed into a text call stream by voice recognition software on both the consumer and the client sides of the call stream.
- the current transcribed call stream is aggregated with at least one other call stream into one or more consumer entities. Patterns may be extracted from a call stream and/or across a plurality of call streams to indicate noteworthy activity and higher level conclusions may be drawn based on the appearance of one or more specific tags.
- the transcribed call stream, caller ID, and extracted pattern data are presented to a client in an e-mail-like “telephone call inbox.”
- a first telephone call inbox screen lists each call per line, which includes a caller ID string, key phrases and/or extracted patterns that are relevant to the client, followed by a partial transcription of the call.
- the client is presented with a second screen that includes a listing of the consumer, the client, the key word(s) or conclusion as a heading or subject, and a line-by-line transcription of the client/consumer conversation, with key words highlighted.
- a computer-implemented method for automatically extracting and presenting transcriptions of telephone calls from at least one consumer to a client comprising the steps of: extracting the identity of a caller from a call received by the client; transcribing the call; aggregating the identity of the caller into at least one consumer entity; extracting at least one pattern from the transcribed call; and distributing the at least one consumer entity and the at least one pattern to the client.
- the at least one consumer entity may a single consumer entity and the extracted identity is aggregated with at least one other identity of the same caller. At least one of the extracted identity and the other identity is based on at least one of a caller ID, a telephone number, and an identity found in the transcribed call.
- the call may be transcribed into at least a text-based call steam.
- Aggregating the identity of the caller into at least one consumer entity may further comprise aggregating a telephone number associated with the call with other telephone numbers associated with the at least one consumer such that multiple call streams are grouped as a single consumer entity.
- the at least one pattern may be extracted from at least one call to draw higher level conclusions to apply a tag to the at least one pattern based on the appearance of at least one of a word and phrase.
- the pattern may be extracted based on an industry-specific pattern set.
- the industry-specific pattern set may be relevant to a business associated with the client.
- the state of a call may be updated based on at least one of the at least one pattern and higher level conclusions.
- a content-based fraud detector rule set may be applied against the transcribed call to detect fraudulent calls.
- at least one of a telephone number and a caller ID associated with the call may be sent over the Internet to a Web site which maintains a record of customer complaints associated with the telephone number to detect fraudulent calls.
- the call may be transcribed into a text call stream on both the consumer and the client sides of the call along with an audio waveform of the call, one channel of the audio waveform being associated with the consumer and the other channel being associated with the client.
- the call may be transcribed using voice recognition software.
- Data append services may be applied to at least one of a caller ID associated with the call and at least one phone number recognized during transcription of the call.
- distributing the at least one consumer entity and the at least one pattern to the client may further comprise distributing the transcribed call, caller ID, and extracted pattern data to the client in a formatted call log.
- the formatted call log may have the format of an e-mail-like telephone call inbox.
- a partial transcription of the call may be provided with the at least one pattern highlighted.
- a Web server may receive a double clicks on the partial transcription from the client, causing a second screen to be transmitted to the client that includes a listing of the consumer entity, the client, the at least one pattern as a heading or subject, and a line-by-line transcription of a client/consumer conversation with key words highlighted.
- FIG. 1 is a process flow diagram illustrating exemplary steps for extracting and presenting transcriptions of call streams from consumers to a client in an e-mail-like telephone call inbox, according to an embodiment of the present invention
- FIG. 2 depicts a system for extracting and presenting transcriptions of call streams from consumers to a client in an e-mail-like telephone call inbox, according to an embodiment of the present invention
- FIG. 3 is a screen shot of a formatted telephone call inbox web page, according to an embodiment of the present invention.
- FIG. 4 is a screen shot of a second telephone call inbox web page displayed when a client double clicks on a specific call stream of the formatted telephone call inbox screen displayed in FIG. 3 ;
- FIG. 5A shows a consumer database table that lists the names and addresses of consumers associated with a client, according to an embodiment of the present invention
- FIG. 5B shows a consumer to call match database table which maps call streams to consumers, according to an embodiment of the present invention
- FIG. 5C shows a call database table that lists a high level summary of call streams to the client, according to an embodiment of the present invention
- FIG. 5D shows a call data storage database table that lists low level details a single call stream, according to an embodiment of the present invention.
- FIG. 6 is a process flow diagram illustrating the exemplary steps of FIG. 1 in greater detail, according to an embodiment of the present invention.
- FIG. 1 is a process flow diagram illustrating exemplary steps for extracting and presenting transcriptions of call streams from consumers to a client in an e-mail-like telephone call inbox, according to an embodiment of the present invention.
- the identity of at least one consumer is extracted from a phone call.
- non-consumer fraudulent call data is automatically filtered out of the call.
- the call is transcribed into a text call stream on both the consumer and the client sides of the call.
- the audio waveform of a call stream is stored, with one (e.g., the left) channel being associated with the consumer and the other (e.g., the right) channel being associated with the client.
- Each of the channels is transcribed separately with timing data stored.
- a client may receive more than one telephone call from the same consumer, but the consumer may be associated with more than one telephone number.
- the consumer may call from their business telephone, their home telephone, their cell phone, etc. Therefore, at step 14 , the current call stream may be aggregated with at least one other call stream into one or more consumer entities.
- Information on the identification of consumer entities is extracted and linked to a database of consumers. In this way, multiple call streams may be grouped as a single consumer entity in a data model. The aggregation of phone calls from a single consumer with different telephone numbers provides the client with a better understanding of the number of consumer entities with which the client is associated.
- patterns may be extracted from a call stream and/or across a plurality of call streams to indicate noteworthy activity and to draw higher level conclusions based on the appearance of one or more specific tags. For example, mentioning payment and the word ‘appointment’ may indicate the conclusion that an appointment was booked on a call.
- data append services may be applied to the caller ID associated with a call or to one or more phone numbers recognized during transcription of a call stream and linked and appended to a data record associated with the call stream. Data collected from a data append service may include the first and last name and billing address of the telephone number captured.
- the transcribed call stream, caller ID, and extracted pattern data are presented to a client in an e-mail-like “telephone call inbox.”
- FIG. 2 depicts a system 30 for extracting and presenting transcriptions of call streams from consumers to a client in an e-mail-like telephone call inbox, according to an embodiment of the present invention.
- the system 30 includes a telephony server 32 , the telephony server 32 being associated with a data store 34 , an analysis server 36 , a web server 38 , and a data entry terminal/computer 40 .
- the telephony server 32 may be a traditional private branch exchange (PBX) or a voice-over-IP (VOIP) server.
- PBX private branch exchange
- VOIP voice-over-IP
- the telephony server 32 may be configured for intercepting calls from consumers 42 a - 42 n via the public switched telephone network (PSTN) 44 , the Internet 46 , or both; for completing those calls to clients 48 a - 48 n over the PSTN 44 , the Internet 46 , or both; for bridging calls between the consumers 42 a - 42 n and the clients 48 a - 48 n , for extracting caller IDs belonging to the consumers 42 a - 42 n , for recording telephone calls received by the clients 48 a - 48 n ; and for logging the recorded calls and the caller IDs to the data store 34 .
- PSTN public switched telephone network
- the analysis server 36 may be configured for transcribing received telephone calls into voice-recognized text call streams and converting caller IDs to physical names and addresses of the consumers 42 a - 42 n either from information associated with the PSTN 44 , information extracted from web sites located on the Internet 46 and/or the telephony server 32 , or information extracted from the call streams.
- the analysis server 36 may be further configured for analyzing and extracting patterns within and among the call streams to produce key words or phrases that are relevant to a client 48 a - 48 n , such as completing the pattern tagging of calls and aggregation of calls into consumer entities as described in steps 14 - 18 of FIG. 1 above.
- the data store 34 stores a number of tables needed for analysis and for storing and the retrieval of call streams to be described in connection with FIGS. 5A-5D hereinbelow.
- a domain analyst may enter typical key word patterns to be extracted from calls manually via the data entry terminal/computer 40 .
- key word patterns may be provided by third party software.
- the web server 38 is configured for presenting analyzed call streams comprising grouped consumers to the clients 48 a - 48 n in the form of formatted call logs and telephone call inboxes to be described hereinbelow in connection with FIGS. 3 and 4 , respectively.
- the telephony server 32 , the analysis server 36 , and the web server 38 may each comprise at least one processor, which may be included in a personal computer, a work station, a mainframe computer, or any other device having enough processing power for carrying out the present invention.
- one or more of the servers/computers 32 , 36 , 38 may be aggregated into one or more processors running on a single personal computer, a work station, or a mainframe computer.
- Each of the clients 48 a - 48 n may have a voice line or VOIP telephone 48 a - 48 n configured to receive inbound calls via the telephony server 32 .
- Each of the clients 48 a - 48 n may have one or more terminals/personal computer/workstations 50 a - 50 n for logging into the web server 38 in order to view their telephone call inbox.
- the client 48 a may communicate with a consumers 42 a over the PSTN 44 if the telephony server 32 is a PBX, the client 48 a have a voice line phone, and the consumer 48 a has a voice line phone; over the Internet 46 and the PSTN 44 if the telephony server 32 is a VOIP server, the client 48 n has a VOIP phone, and the consumer 42 a has a voice line phone; and, over the Internet 46 if the telephony server 32 is a VOIP server and the client 48 n has a VOW phone.
- FIG. 3 is a screen shot of a formatted telephone call inbox 52 , according to an embodiment of the present invention.
- a client 48 a enters a URL corresponding to the Web server 38
- the client 48 a is presented with a user login screen (not shown) which prompts the client 48 a to enter a login and password.
- the client 48 a is directed to the formatted telephone call inbox 52 assigned to the client 48 a , which displays call listings from the consumers 42 a - 42 n .
- the formatted telephone call inbox 52 contains a plurality of call entries 54 .
- Each of the call entries 54 includes a consumer name field 56 , which may be a personal name, a company name, or a telephone number.
- the consumer name field 56 may be extracted either from the caller ID, a third party name service, or from the call stream itself and is described in greater detail hereinbelow in connection with FIG. 6 .
- Each of the call entries 54 also includes key words or tags 60 , which may include words or phrases of importance extracted from the call stream and which are significant to the business of the client 48 a to be described in greater detail hereinbelow in connection with FIG. 6 .
- Key words or tags 62 may also indicate a wrong or non-legitimate telephone number/call.
- a freeform transcription search box 64 permits filtering of the call entries 54 so as to display only those call entries 54 that contain entered key word(s) or consumer name fields 56 .
- FIG. 4 is a screen shot of a second call screen 68 that is displayed when the client 48 a double clicks on a specific call entry 54 in the formatted telephone call inbox 52 of FIG. 3 .
- the call screen 68 displays many of the same types of fields common to e-mail programs.
- a field 70 is populated by the consumer name field 56 , which may also include the calling telephone number.
- a field 72 is populated by the name of the client 48 a and a telephone number.
- the message body 74 includes the transcription of the specific call, with alternating portions 76 , 78 being a transcription of the client and consumer portions of the call stream, respectively. Key words or phrases 80 may be highlighted in the call stream
- FIGS. 5A-5D show examples of tables located in the data store 34 for processing calls and the telephone call inboxes, according to an embodiment of the present invention.
- a consumer database table 90 includes a plurality of records 92 .
- Each of the records 92 corresponds to one of the consumers 42 a - 42 n associated with a client 48 a - 48 n and is derived from incoming phone calls.
- Each of the records 92 contains fields for at least an arbitrarily assigned consumer number 93 , the first name 94 , the last name 96 , the street address 98 , and the telephone number 100 associated with a consumer 42 a - 42 n , the street address being derived by methods to be described in connection with FIG. 6 hereinbelow.
- the same client may have more than one record 92 corresponding to different telephone numbers 100 and street addresses 98 .
- a consumer to call match database table 102 maps call streams 104 to consumers 106 .
- the designation for a call stream 104 is a numerical value assigned in increasing order as calls are received.
- the designation for a consumer 106 corresponds to the consumer number 93 of FIG. 5A .
- a call database table 108 lists a high level summary of a call stream.
- Each record 110 in the call database table 108 includes at least a call ID number 112 corresponding to the designation for a call stream 104 of FIG. 5B , a “From” field 114 indicating the telephone number of one of the consumers 42 a - 42 n or an invalid caller ID, a “To” field 116 corresponding to one of the clients 48 a - 48 n , and a “length” field 118 indicating the length of time of the call corresponding to the call stream.
- a call data storage database table 120 lists low level details of each call stream associated with each client.
- the call data storage database table 120 lists, for each call stream, information concerning key words and phrases and the extracted pattern found in the call stream along with time stamps and pattern types.
- Each record includes at least a call ID number 124 corresponding to the designation for a call stream 104 of FIG. 5B , a type field 126 listing the type of extracted key word or phrase, an offset field 128 indicating the relative time position in the call stream at which the key word or conclusion occurred, a length field 130 indicating the time length corresponding to the length of time needed to utter the key word or phrase, and a data field 132 listing the key word, phrase, or conclusion.
- FIG. 6 is a process flow diagram illustrating the exemplary steps of FIG. 1 in greater detail.
- FIG. 2 indicates several scenarios for interaction between the consumers 42 a - 42 n and the clients 48 a - 48 n , for illustrative purposes, the following example shows interactions in which a client 48 a has a VOIP telephone for communicating with a consumer 42 a that has voice line phone.
- a consumer 42 a places a call to a client 48 a .
- the PSTN 44 and Internet 46 route the call to the telephony server 32 .
- the PSTN 44 via the Internet 46 rings a telephone number assigned to the telephony server 32 on behalf of the client 48 a .
- the telephony server 32 completes the call to the client 48 a over the Internet 46 , bridges the call, extracts the caller ID belonging to the consumer 42 a , and immediately begins recording the call even before the client 48 a answers their telephone.
- Immediate recording is one method for detecting fraudulent calls, as a recoded telemarketing call may begin playing as soon as a receiving telephone is rung.
- the call is recorded on two channels, with one (e.g., the left) channel being associated with the consumer 42 a and the other (e.g., the right) channel being associated with the client 48 a.
- fraud filters are applied to the call.
- the fraud filters may match the extracted caller ID against a database of known telemarketers. Additional rules are applied to detect fraudulent calls via phrases uttered in the call stream associated with the transcribed call to be discussed hereinbelow, or a signal indicating a fax machine. If, at step 148 , the fraud filters indicate a fraudulent call, then further processing terminates.
- the consumer database table 90 of FIG. 5A is searched for an existing consumer record. All of the fields of a record need to match an existing consumer.
- a first name, last name, and street address are derived from the caller ID of an incoming call based on techniques to be described hereinbelow and compared to the first name 94 , the last name 96 , and the street address 98 of each record 92 in the consumer database table 90 .
- the name and street address information derived from the caller ID of the incoming call may be derived from a number of sources.
- CNAM Calling Name Delivery
- Other methods for obtaining name and address information are by consulting third party services such as Acxiom, InfoUSA and/or Accudata.
- Still other methods may include matching census collected common first and last names in a database available on the Internet 46 to a name that is partially truncated in the CNAM. If no name and address is matched, then the telephone number of the consumer 48 a is displayed in the telephone call inbox depicted in FIGS. 3 and 4 .
- step 154 If, at step 154 , an existing consumer record is found to match in the consumer database table 90 , then at step 155 , the call is linked to the consumer identified in the consumer record found. If, at step 154 , no existing consumer record is found to match the data in the incoming call, then at step 156 , a new consumer entity is created and stored in the consumer database table 90 , with the consumer linked to this record. At step 157 , data append services are run for the new consumer. At step 158 , for either an existing consumer or a new consumer, a new record is created in each of tables 102 , 108 , and 120 for the new call.
- a transcription of the call between the client 48 a and the consumer 42 a is produced using voice recognition software and stored in the data store 34 along with the recording of the call, preferably in MP3 format.
- the call transcription is analyzed in real time or near real time to detect patterns using an industry-specific pattern set for the purpose of deriving the tags described above.
- the pattern set against which portions of the call are matched is provided by third party software or manually by a domain expert.
- the industry-specific set of patterns is relevant to the business associated with the client 48 a , such as information pertaining to chiropractors or automotive repair shops, etc.
- Some patterns may be extracted that are independent of an industry, such as the consumer's telephone number, which may be then be substituted in the consumer database table 90 for an existing telephone number if the consumer 42 a called from a general number and the phone number mentioned in the call is a direct number (e.g., 973-515-2143 substituted for 973-515-2000).
- a content-based fraud detector rule set may be run against the transcribed call stream to further detect fraudulent calls, including patterns detected such as the word “sorry” or “wrong number.” If the time codes of the call recorded in table 120 indicate that during, say, the first 15 seconds of the call, the consumer 42 a speaks continuously without the client 48 a being allowed to speak, then this may be interpreted by the fraud detector as a sign of a telemarketer. Other pattern phrases in the call may be detected as fraudulent, such as “looking for a way out of credit card debt.” Alternatively, the telephone number or caller ID may be sent over the Internet 46 to a web site which maintains a record of customer complaints associated with the telephone number, which may be employed as another means of detecting fraudulent telephone calls. If, at step 165 , the fraud filters indicate a fraudulent call, then at step 166 , a tag indicating a fraudulent call is associated with the transcription of the call in the data storage database table 120 .
- the fraud filters indicate a valid call
- the telephone number, name, and address are linked to a new or existing consumer entity in the consumer database table 90 .
- the telephone number of the call stream is aggregated with other numbers associated with the same consumer 42 a in the data store 34 such that multiple call streams may be grouped as a single consumer entity in the tables 90 , 102 , 108 , of FIGS. 5A-5C , respectively. This permits aggregate information to be determined and linked (e.g., ‘what is the current status of this caller to this single client?’, ‘what is the current status of this caller across all clients?’).
- the linkage is typically created by linking a name and address of the consumer 42 a in the consumer database table 90 to one or more phone numbers found in a caller ID or spoken aloud and extracted from the call transcription.
- a single name entity is displayed in the telephone call inbox of FIGS. 3 and 4 for all telephone numbers associated with the name entity.
- an industry specific rule set is applied to the telephone call to draw conclusions concerning the patterns in the call stream.
- the industry specific rule set is provided by third party software or manually by a domain expert.
- a sample conclusion that may be drawn from a call stream is that the client 42 a wishes to make an “appointment” to meet with the client 48 a .
- Another type of conclusion may be drawn by comparing patterns in a current call stream to patterns in other calls streams stored in the data store 34 from the same client 42 a or across all clients 42 a - 42 n , so as to draw conclusions concerning, for example, total call volume per month for a given consumer 42 a .
- the pattern words or conclusions may be appended to the beginning of the transcription of the call stream in BOLD in the call entry 54 for the call as depicted in FIGS. 3 and 4 and added as “type” and/or “data” entries in the call data storage database table 120 of FIG. 5D .
- a conclusion may be made in a “type” field for call number 2 for client 48 a and consumer 42 a which states that “an appointment was booked.”
- the “state” or “status” of the consumer 42 a may be updated.
- the state of calls for a consumer 42 a may be regarded as passing through stages of a buying cycle.
- the first state may be reached after receiving a first call having a pattern that indicated “wanting to book an appointment” with the client 48 a .
- the second state may be reached when the pattern “doctor available” is detected.
- the third state may be reached when the pattern “appointment booked” is detected.
- the fourth state may be reached when the pattern “reschedule an appointment” is detected.
- the final state may be reached and noted in the data store 34 when the pattern “cancelled appointment” is detected.
- a new process begins when, at step 174 , the client logs in to the web site corresponding to the system 30 using their computer 50 a .
- the transcribed call stream, consumer name, and pattern data are extracted from tables 90 , 102 , 108 , and 120 and other data located in the data store 34 are presented to a client in the form of the e-mail-like telephone call inbox screens 52 , 68 described above in connection with FIGS. 3 and 4 , respectively.
- FIGS. 3 and 4 described above in connection with FIGS. 3 and 4
- the telephone inbox screens 52 , 68 provide advantages over existing transcription services.
- the client gains a deeper understanding of their consumer base to allow the client to plan inventories, track their business volume and revenues, etc.
- the client can filter invalid consumers from valid consumers.
- a client can better instruct executive assistants as to which consumers are of higher versus lower priority.
- the time needed for reviewing call transcriptions is greatly reduced. In this way, telephone call inboxes provides the client with a tool for increasing efficiency and reducing operating costs.
Abstract
Description
- This application claims the benefit of U.S. provisional patent application No. 61/229,820 filed Jul. 30, 2009, the disclosure of which is incorporated herein by reference in its entirety.
- The present invention relates generally to telecommunications, and more particularly, to a method and system for creating an e-mail-like telephone call inbox of transcribed telephone calls between consumers and a business.
- When a business (hereinafter, the “client(s)”) uses pay per call advertising, the client may receive a number of calls from consumers seeking their products or services, etc. When business is good, the number of calls may be overwhelming and too time intensive to monitor. What is desirable to monitor includes determining the number of consumers and extracting significant data from consumer interactions, such as the number and type of products sold, the state or phase of a complicated transaction, or the type of consumer that is interested in a particular product or service.
- To keep track of telephone calls and their content (hereinafter, the “consumer(s)”), the client may subscribe to an audio recording and or transcription service. Unfortunately, if the client receives, for example, 30 phone calls a day, and playing the recorded phone messages takes on average five to ten minutes apiece (or a similar amount of time in transcribed form) it may take three hours to listen/read all of the calls, which becomes tedious. Moreover, the client may receive only the caller ID number and a short string identifying the name of the consumer on their telephone, which provides little information about the consumer. The client also needs to manually and mentally extract relevant information from each call, yet may have no tools for finding patterns in the calls that may affect their business.
- Accordingly, what would be desirable, but has not yet been provided, is and method and system that simplifies the management of call streams from consumers to a client. The method and system would be capable of providing useful data extracted from the call streams that is presented to the client in a user-friendly form.
- The above-described problems are addressed and a technical solution is achieved in the art by providing a method and system for extracting useful data from calls streams of consumers received by a client and presenting the data to the client in an e-mail-like “telephone call inbox.” As used herein, the term “client” may refer to any person or organization that may employ the method and/or system of the present invention, which may include, but is not limited to, an individual, a non-profit organization such as a university, and a for-profit business. As used herein, the term “consumer” may refer to any person or organization that may call into the system using the method of the present invention, which may include, but is not limited to, an individual, a non-profit organization such as a university, and a for-profit business. When a call is received from a potential consumer at a telephony server, the telephony server completes the call to a client, bridges the call, and begins recording the call. The identity of at least one consumer calling a client is extracted from the call. As part of the identification process, non-consumer fraudulent call data is automatically filtered out of the call (stream). A call is transcribed into a text call stream by voice recognition software on both the consumer and the client sides of the call stream. The current transcribed call stream is aggregated with at least one other call stream into one or more consumer entities. Patterns may be extracted from a call stream and/or across a plurality of call streams to indicate noteworthy activity and higher level conclusions may be drawn based on the appearance of one or more specific tags.
- The transcribed call stream, caller ID, and extracted pattern data are presented to a client in an e-mail-like “telephone call inbox.” A first telephone call inbox screen lists each call per line, which includes a caller ID string, key phrases and/or extracted patterns that are relevant to the client, followed by a partial transcription of the call. To view the entire transcription; the client double clicks on the partial transcription. The client is presented with a second screen that includes a listing of the consumer, the client, the key word(s) or conclusion as a heading or subject, and a line-by-line transcription of the client/consumer conversation, with key words highlighted.
- According to an embodiment of the present invention, a computer-implemented method for automatically extracting and presenting transcriptions of telephone calls from at least one consumer to a client is disclosed, comprising the steps of: extracting the identity of a caller from a call received by the client; transcribing the call; aggregating the identity of the caller into at least one consumer entity; extracting at least one pattern from the transcribed call; and distributing the at least one consumer entity and the at least one pattern to the client. The at least one consumer entity may a single consumer entity and the extracted identity is aggregated with at least one other identity of the same caller. At least one of the extracted identity and the other identity is based on at least one of a caller ID, a telephone number, and an identity found in the transcribed call. The call may be transcribed into at least a text-based call steam. Aggregating the identity of the caller into at least one consumer entity may further comprise aggregating a telephone number associated with the call with other telephone numbers associated with the at least one consumer such that multiple call streams are grouped as a single consumer entity.
- According to an embodiment of the present invention, the at least one pattern may be extracted from at least one call to draw higher level conclusions to apply a tag to the at least one pattern based on the appearance of at least one of a word and phrase. The pattern may be extracted based on an industry-specific pattern set. The industry-specific pattern set may be relevant to a business associated with the client. The state of a call may be updated based on at least one of the at least one pattern and higher level conclusions.
- According to an embodiment of the present invention, a content-based fraud detector rule set may be applied against the transcribed call to detect fraudulent calls. Alternatively, at least one of a telephone number and a caller ID associated with the call may be sent over the Internet to a Web site which maintains a record of customer complaints associated with the telephone number to detect fraudulent calls.
- According to an embodiment of the present invention, the call may be transcribed into a text call stream on both the consumer and the client sides of the call along with an audio waveform of the call, one channel of the audio waveform being associated with the consumer and the other channel being associated with the client. The call may be transcribed using voice recognition software.
- Data append services may be applied to at least one of a caller ID associated with the call and at least one phone number recognized during transcription of the call.
- According to an embodiment of the present invention, distributing the at least one consumer entity and the at least one pattern to the client may further comprise distributing the transcribed call, caller ID, and extracted pattern data to the client in a formatted call log. The formatted call log may have the format of an e-mail-like telephone call inbox. A partial transcription of the call may be provided with the at least one pattern highlighted.
- A Web server may receive a double clicks on the partial transcription from the client, causing a second screen to be transmitted to the client that includes a listing of the consumer entity, the client, the at least one pattern as a heading or subject, and a line-by-line transcription of a client/consumer conversation with key words highlighted.
- The present invention will be more readily understood from the detailed description of an exemplary embodiment presented below considered in conjunction with the attached drawings, of which:
-
FIG. 1 is a process flow diagram illustrating exemplary steps for extracting and presenting transcriptions of call streams from consumers to a client in an e-mail-like telephone call inbox, according to an embodiment of the present invention; -
FIG. 2 depicts a system for extracting and presenting transcriptions of call streams from consumers to a client in an e-mail-like telephone call inbox, according to an embodiment of the present invention; -
FIG. 3 is a screen shot of a formatted telephone call inbox web page, according to an embodiment of the present invention; -
FIG. 4 is a screen shot of a second telephone call inbox web page displayed when a client double clicks on a specific call stream of the formatted telephone call inbox screen displayed inFIG. 3 ; -
FIG. 5A shows a consumer database table that lists the names and addresses of consumers associated with a client, according to an embodiment of the present invention; -
FIG. 5B shows a consumer to call match database table which maps call streams to consumers, according to an embodiment of the present invention; -
FIG. 5C shows a call database table that lists a high level summary of call streams to the client, according to an embodiment of the present invention; -
FIG. 5D shows a call data storage database table that lists low level details a single call stream, according to an embodiment of the present invention; and -
FIG. 6 is a process flow diagram illustrating the exemplary steps ofFIG. 1 in greater detail, according to an embodiment of the present invention. - It is to be understood that the attached drawings are for purposes of illustrating the concepts of the invention and may not be to scale.
-
FIG. 1 is a process flow diagram illustrating exemplary steps for extracting and presenting transcriptions of call streams from consumers to a client in an e-mail-like telephone call inbox, according to an embodiment of the present invention. Atstep 10, the identity of at least one consumer is extracted from a phone call. As part of the identification process, non-consumer fraudulent call data is automatically filtered out of the call. Atstep 12, the call is transcribed into a text call stream on both the consumer and the client sides of the call. The audio waveform of a call stream is stored, with one (e.g., the left) channel being associated with the consumer and the other (e.g., the right) channel being associated with the client. Each of the channels is transcribed separately with timing data stored. A client may receive more than one telephone call from the same consumer, but the consumer may be associated with more than one telephone number. For example, the consumer may call from their business telephone, their home telephone, their cell phone, etc. Therefore, atstep 14, the current call stream may be aggregated with at least one other call stream into one or more consumer entities. Information on the identification of consumer entities is extracted and linked to a database of consumers. In this way, multiple call streams may be grouped as a single consumer entity in a data model. The aggregation of phone calls from a single consumer with different telephone numbers provides the client with a better understanding of the number of consumer entities with which the client is associated. Atstep 16, patterns ('tags') may be extracted from a call stream and/or across a plurality of call streams to indicate noteworthy activity and to draw higher level conclusions based on the appearance of one or more specific tags. For example, mentioning payment and the word ‘appointment’ may indicate the conclusion that an appointment was booked on a call. Atstep 18, data append services may be applied to the caller ID associated with a call or to one or more phone numbers recognized during transcription of a call stream and linked and appended to a data record associated with the call stream. Data collected from a data append service may include the first and last name and billing address of the telephone number captured. Atstep 20, the transcribed call stream, caller ID, and extracted pattern data are presented to a client in an e-mail-like “telephone call inbox.” -
FIG. 2 depicts asystem 30 for extracting and presenting transcriptions of call streams from consumers to a client in an e-mail-like telephone call inbox, according to an embodiment of the present invention. Thesystem 30 includes atelephony server 32, thetelephony server 32 being associated with adata store 34, ananalysis server 36, aweb server 38, and a data entry terminal/computer 40. Thetelephony server 32 may be a traditional private branch exchange (PBX) or a voice-over-IP (VOIP) server. Thetelephony server 32 may be configured for intercepting calls from consumers 42 a-42 n via the public switched telephone network (PSTN) 44, theInternet 46, or both; for completing those calls to clients 48 a-48 n over thePSTN 44, theInternet 46, or both; for bridging calls between the consumers 42 a-42 n and the clients 48 a-48 n, for extracting caller IDs belonging to the consumers 42 a-42 n, for recording telephone calls received by the clients 48 a-48 n; and for logging the recorded calls and the caller IDs to thedata store 34. Theanalysis server 36 may be configured for transcribing received telephone calls into voice-recognized text call streams and converting caller IDs to physical names and addresses of the consumers 42 a-42 n either from information associated with thePSTN 44, information extracted from web sites located on theInternet 46 and/or thetelephony server 32, or information extracted from the call streams. Theanalysis server 36 may be further configured for analyzing and extracting patterns within and among the call streams to produce key words or phrases that are relevant to a client 48 a-48 n, such as completing the pattern tagging of calls and aggregation of calls into consumer entities as described in steps 14-18 ofFIG. 1 above. Thedata store 34 stores a number of tables needed for analysis and for storing and the retrieval of call streams to be described in connection withFIGS. 5A-5D hereinbelow. A domain analyst may enter typical key word patterns to be extracted from calls manually via the data entry terminal/computer 40. Alternatively, key word patterns may be provided by third party software. Theweb server 38 is configured for presenting analyzed call streams comprising grouped consumers to the clients 48 a-48 n in the form of formatted call logs and telephone call inboxes to be described hereinbelow in connection withFIGS. 3 and 4 , respectively. Thetelephony server 32, theanalysis server 36, and theweb server 38 may each comprise at least one processor, which may be included in a personal computer, a work station, a mainframe computer, or any other device having enough processing power for carrying out the present invention. Alternatively, one or more of the servers/computers - Each of the clients 48 a-48 n may have a voice line or VOIP telephone 48 a-48 n configured to receive inbound calls via the
telephony server 32. Each of the clients 48 a-48 n may have one or more terminals/personal computer/workstations 50 a-50 n for logging into theweb server 38 in order to view their telephone call inbox. The client 48 a may communicate with aconsumers 42 a over thePSTN 44 if thetelephony server 32 is a PBX, the client 48 a have a voice line phone, and the consumer 48 a has a voice line phone; over theInternet 46 and thePSTN 44 if thetelephony server 32 is a VOIP server, theclient 48 n has a VOIP phone, and theconsumer 42 a has a voice line phone; and, over theInternet 46 if thetelephony server 32 is a VOIP server and theclient 48 n has a VOW phone. -
FIG. 3 is a screen shot of a formattedtelephone call inbox 52, according to an embodiment of the present invention. When a client 48 a enters a URL corresponding to theWeb server 38, the client 48 a is presented with a user login screen (not shown) which prompts the client 48 a to enter a login and password. Once logged in, the client 48 a is directed to the formattedtelephone call inbox 52 assigned to the client 48 a, which displays call listings from the consumers 42 a-42 n. The formattedtelephone call inbox 52 contains a plurality ofcall entries 54. Each of thecall entries 54 includes aconsumer name field 56, which may be a personal name, a company name, or a telephone number. Theconsumer name field 56 may be extracted either from the caller ID, a third party name service, or from the call stream itself and is described in greater detail hereinbelow in connection withFIG. 6 . Each of thecall entries 54 also includes key words ortags 60, which may include words or phrases of importance extracted from the call stream and which are significant to the business of the client 48 a to be described in greater detail hereinbelow in connection withFIG. 6 . Key words ortags 62 may also indicate a wrong or non-legitimate telephone number/call. A freeformtranscription search box 64 permits filtering of thecall entries 54 so as to display only thosecall entries 54 that contain entered key word(s) or consumer name fields 56. -
FIG. 4 is a screen shot of asecond call screen 68 that is displayed when the client 48 a double clicks on aspecific call entry 54 in the formattedtelephone call inbox 52 ofFIG. 3 . Referring now toFIGS. 3 and 4 , thecall screen 68 displays many of the same types of fields common to e-mail programs. Afield 70 is populated by theconsumer name field 56, which may also include the calling telephone number. Afield 72 is populated by the name of the client 48 a and a telephone number. Themessage body 74 includes the transcription of the specific call, with alternatingportions 76, 78 being a transcription of the client and consumer portions of the call stream, respectively. Key words orphrases 80 may be highlighted in the call stream -
FIGS. 5A-5D show examples of tables located in thedata store 34 for processing calls and the telephone call inboxes, according to an embodiment of the present invention. InFIG. 5A , a consumer database table 90 includes a plurality ofrecords 92. Each of therecords 92 corresponds to one of the consumers 42 a-42 n associated with a client 48 a-48 n and is derived from incoming phone calls. Each of therecords 92 contains fields for at least an arbitrarily assignedconsumer number 93, thefirst name 94, thelast name 96, thestreet address 98, and thetelephone number 100 associated with a consumer 42 a-42 n, the street address being derived by methods to be described in connection withFIG. 6 hereinbelow. It should be noted that the same client may have more than onerecord 92 corresponding todifferent telephone numbers 100 and street addresses 98. - In
FIG. 5B , a consumer to call match database table 102 maps callstreams 104 toconsumers 106. The designation for acall stream 104 is a numerical value assigned in increasing order as calls are received. The designation for aconsumer 106 corresponds to theconsumer number 93 ofFIG. 5A . As shown inFIG. 5B , there may beseveral call streams 104 associated with thesame consumer 106. - In
FIG. 5C , a call database table 108 lists a high level summary of a call stream. Eachrecord 110 in the call database table 108 includes at least acall ID number 112 corresponding to the designation for acall stream 104 ofFIG. 5B , a “From”field 114 indicating the telephone number of one of the consumers 42 a-42 n or an invalid caller ID, a “To”field 116 corresponding to one of the clients 48 a-48 n, and a “length”field 118 indicating the length of time of the call corresponding to the call stream. - In
FIG. 5D , a call data storage database table 120 lists low level details of each call stream associated with each client. The call data storage database table 120 lists, for each call stream, information concerning key words and phrases and the extracted pattern found in the call stream along with time stamps and pattern types. Each record includes at least acall ID number 124 corresponding to the designation for acall stream 104 ofFIG. 5B , atype field 126 listing the type of extracted key word or phrase, an offsetfield 128 indicating the relative time position in the call stream at which the key word or conclusion occurred, alength field 130 indicating the time length corresponding to the length of time needed to utter the key word or phrase, and adata field 132 listing the key word, phrase, or conclusion. -
FIG. 6 is a process flow diagram illustrating the exemplary steps ofFIG. 1 in greater detail. AlthoughFIG. 2 indicates several scenarios for interaction between the consumers 42 a-42 n and the clients 48 a-48 n, for illustrative purposes, the following example shows interactions in which a client 48 a has a VOIP telephone for communicating with aconsumer 42 a that has voice line phone. - Referring now to
FIGS. 1 , 2, 5A-5D, and 6, atstep 140, aconsumer 42 a places a call to a client 48 a. Atstep 142, thePSTN 44 andInternet 46 route the call to thetelephony server 32. According to an embodiment of the present invention, thePSTN 44 via theInternet 46 rings a telephone number assigned to thetelephony server 32 on behalf of the client 48 a. As described above, atstep 144, thetelephony server 32 completes the call to the client 48 a over theInternet 46, bridges the call, extracts the caller ID belonging to theconsumer 42 a, and immediately begins recording the call even before the client 48 a answers their telephone. Immediate recording is one method for detecting fraudulent calls, as a recoded telemarketing call may begin playing as soon as a receiving telephone is rung. The call is recorded on two channels, with one (e.g., the left) channel being associated with theconsumer 42 a and the other (e.g., the right) channel being associated with the client 48 a. - At
step 146, fraud filters are applied to the call. The fraud filters may match the extracted caller ID against a database of known telemarketers. Additional rules are applied to detect fraudulent calls via phrases uttered in the call stream associated with the transcribed call to be discussed hereinbelow, or a signal indicating a fax machine. If, atstep 148, the fraud filters indicate a fraudulent call, then further processing terminates. - If, at
step 148, the call is deemed to be legitimate, then atstep 152, the consumer database table 90 ofFIG. 5A is searched for an existing consumer record. All of the fields of a record need to match an existing consumer. A first name, last name, and street address are derived from the caller ID of an incoming call based on techniques to be described hereinbelow and compared to thefirst name 94, thelast name 96, and thestreet address 98 of each record 92 in the consumer database table 90. The name and street address information derived from the caller ID of the incoming call may be derived from a number of sources. One source is Calling Name Delivery (CNAM) associated with a call that is located within the in-band signaling information within the call, which is a 15 digit name string. Other methods for obtaining name and address information are by consulting third party services such as Acxiom, InfoUSA and/or Accudata. Still other methods may include matching census collected common first and last names in a database available on theInternet 46 to a name that is partially truncated in the CNAM. If no name and address is matched, then the telephone number of the consumer 48 a is displayed in the telephone call inbox depicted inFIGS. 3 and 4 . - If, at
step 154, an existing consumer record is found to match in the consumer database table 90, then atstep 155, the call is linked to the consumer identified in the consumer record found. If, atstep 154, no existing consumer record is found to match the data in the incoming call, then atstep 156, a new consumer entity is created and stored in the consumer database table 90, with the consumer linked to this record. Atstep 157, data append services are run for the new consumer. Atstep 158, for either an existing consumer or a new consumer, a new record is created in each of tables 102, 108, and 120 for the new call. - At
step 160, a transcription of the call between the client 48 a and theconsumer 42 a is produced using voice recognition software and stored in thedata store 34 along with the recording of the call, preferably in MP3 format. Atstep 162, the call transcription is analyzed in real time or near real time to detect patterns using an industry-specific pattern set for the purpose of deriving the tags described above. The pattern set against which portions of the call are matched is provided by third party software or manually by a domain expert. The industry-specific set of patterns is relevant to the business associated with the client 48 a, such as information pertaining to chiropractors or automotive repair shops, etc. Some patterns may be extracted that are independent of an industry, such as the consumer's telephone number, which may be then be substituted in the consumer database table 90 for an existing telephone number if theconsumer 42 a called from a general number and the phone number mentioned in the call is a direct number (e.g., 973-515-2143 substituted for 973-515-2000). - At
step 164, a content-based fraud detector rule set may be run against the transcribed call stream to further detect fraudulent calls, including patterns detected such as the word “sorry” or “wrong number.” If the time codes of the call recorded in table 120 indicate that during, say, the first 15 seconds of the call, theconsumer 42 a speaks continuously without the client 48 a being allowed to speak, then this may be interpreted by the fraud detector as a sign of a telemarketer. Other pattern phrases in the call may be detected as fraudulent, such as “looking for a way out of credit card debt.” Alternatively, the telephone number or caller ID may be sent over theInternet 46 to a web site which maintains a record of customer complaints associated with the telephone number, which may be employed as another means of detecting fraudulent telephone calls. If, atstep 165, the fraud filters indicate a fraudulent call, then atstep 166, a tag indicating a fraudulent call is associated with the transcription of the call in the data storage database table 120. - If, at
step 165, the fraud filters indicate a valid call, then atstep 167, the telephone number, name, and address are linked to a new or existing consumer entity in the consumer database table 90. Atstep 168, the telephone number of the call stream is aggregated with other numbers associated with thesame consumer 42 a in thedata store 34 such that multiple call streams may be grouped as a single consumer entity in the tables 90, 102, 108, ofFIGS. 5A-5C , respectively. This permits aggregate information to be determined and linked (e.g., ‘what is the current status of this caller to this single client?’, ‘what is the current status of this caller across all clients?’). The linkage is typically created by linking a name and address of theconsumer 42 a in the consumer database table 90 to one or more phone numbers found in a caller ID or spoken aloud and extracted from the call transcription. As a result, a single name entity is displayed in the telephone call inbox ofFIGS. 3 and 4 for all telephone numbers associated with the name entity. - At
step 170, an industry specific rule set is applied to the telephone call to draw conclusions concerning the patterns in the call stream. The industry specific rule set is provided by third party software or manually by a domain expert. A sample conclusion that may be drawn from a call stream is that theclient 42 a wishes to make an “appointment” to meet with the client 48 a. Another type of conclusion may be drawn by comparing patterns in a current call stream to patterns in other calls streams stored in thedata store 34 from thesame client 42 a or across all clients 42 a-42 n, so as to draw conclusions concerning, for example, total call volume per month for a givenconsumer 42 a. The pattern words or conclusions may be appended to the beginning of the transcription of the call stream in BOLD in thecall entry 54 for the call as depicted inFIGS. 3 and 4 and added as “type” and/or “data” entries in the call data storage database table 120 ofFIG. 5D . For example, a conclusion may be made in a “type” field forcall number 2 for client 48 a andconsumer 42 a which states that “an appointment was booked.” - At
optional step 172, the “state” or “status” of theconsumer 42 a may be updated. For example, the state of calls for aconsumer 42 a may be regarded as passing through stages of a buying cycle. The first state may be reached after receiving a first call having a pattern that indicated “wanting to book an appointment” with the client 48 a. The second state may be reached when the pattern “doctor available” is detected. The third state may be reached when the pattern “appointment booked” is detected. The fourth state may be reached when the pattern “reschedule an appointment” is detected. The final state may be reached and noted in thedata store 34 when the pattern “cancelled appointment” is detected. Thus, a “state” of a single consumer is tracked across a plurality of calls based on conclusions drawn on each call. Processing terminates. - A new process begins when, at
step 174, the client logs in to the web site corresponding to thesystem 30 using theircomputer 50 a. Upon a successful login, atstep 176, the transcribed call stream, consumer name, and pattern data are extracted from tables 90, 102, 108, and 120 and other data located in thedata store 34 are presented to a client in the form of the e-mail-like telephone call inbox screens 52, 68 described above in connection withFIGS. 3 and 4 , respectively. It would be appreciated by those skilled in the art that several programming techniques may be used for presenting the data in the tables 90, 102, 108, 120 in the format of the telephone call inbox screens 52, 68. - The telephone inbox screens 52, 68 provide advantages over existing transcription services. The client gains a deeper understanding of their consumer base to allow the client to plan inventories, track their business volume and revenues, etc. The client can filter invalid consumers from valid consumers. A client can better instruct executive assistants as to which consumers are of higher versus lower priority. The time needed for reviewing call transcriptions is greatly reduced. In this way, telephone call inboxes provides the client with a tool for increasing efficiency and reducing operating costs.
- It is to be understood that the exemplary embodiments are merely illustrative of the invention and that many variations of the above-described embodiments may be devised by one skilled in the art without departing from the scope of the invention. It is therefore intended that all such variations be included within the scope of the following claims and their equivalents.
Claims (25)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US12/726,932 US8774372B2 (en) | 2009-07-30 | 2010-03-18 | Telephone call inbox |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US22982009P | 2009-07-30 | 2009-07-30 | |
US12/726,932 US8774372B2 (en) | 2009-07-30 | 2010-03-18 | Telephone call inbox |
Publications (2)
Publication Number | Publication Date |
---|---|
US20110026689A1 true US20110026689A1 (en) | 2011-02-03 |
US8774372B2 US8774372B2 (en) | 2014-07-08 |
Family
ID=43527005
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US12/726,932 Expired - Fee Related US8774372B2 (en) | 2009-07-30 | 2010-03-18 | Telephone call inbox |
Country Status (1)
Country | Link |
---|---|
US (1) | US8774372B2 (en) |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100303211A1 (en) * | 2005-04-21 | 2010-12-02 | Victrio | Method and system for generating a fraud risk score using telephony channel based audio and non-audio data |
US20120263285A1 (en) * | 2005-04-21 | 2012-10-18 | Anthony Rajakumar | Systems, methods, and media for disambiguating call data to determine fraud |
WO2012160193A1 (en) * | 2011-05-26 | 2012-11-29 | Jajah Ltd. | Voice conversation analysis utilising keywords |
US8793131B2 (en) | 2005-04-21 | 2014-07-29 | Verint Americas Inc. | Systems, methods, and media for determining fraud patterns and creating fraud behavioral models |
US9460722B2 (en) | 2013-07-17 | 2016-10-04 | Verint Systems Ltd. | Blind diarization of recorded calls with arbitrary number of speakers |
US9503571B2 (en) | 2005-04-21 | 2016-11-22 | Verint Americas Inc. | Systems, methods, and media for determining fraud patterns and creating fraud behavioral models |
US9571652B1 (en) | 2005-04-21 | 2017-02-14 | Verint Americas Inc. | Enhanced diarization systems, media and methods of use |
US9875742B2 (en) | 2015-01-26 | 2018-01-23 | Verint Systems Ltd. | Word-level blind diarization of recorded calls with arbitrary number of speakers |
US9875739B2 (en) | 2012-09-07 | 2018-01-23 | Verint Systems Ltd. | Speaker separation in diarization |
US9984706B2 (en) | 2013-08-01 | 2018-05-29 | Verint Systems Ltd. | Voice activity detection using a soft decision mechanism |
US10134401B2 (en) | 2012-11-21 | 2018-11-20 | Verint Systems Ltd. | Diarization using linguistic labeling |
CN108900342A (en) * | 2018-07-03 | 2018-11-27 | 阳光保险集团股份有限公司 | Telephone service processing method, telephone service processing unit and server |
CN109147801A (en) * | 2018-09-30 | 2019-01-04 | 深圳市元征科技股份有限公司 | voice interactive method, system, terminal and storage medium |
US10657971B1 (en) | 2017-12-15 | 2020-05-19 | NortonLifeLock Inc. | Systems and methods for detecting suspicious voice calls |
US20200162509A1 (en) * | 2018-11-19 | 2020-05-21 | OLX Global B.V. | Voice notes fraud detection |
US10887452B2 (en) | 2018-10-25 | 2021-01-05 | Verint Americas Inc. | System architecture for fraud detection |
US10971148B2 (en) * | 2018-03-30 | 2021-04-06 | Honda Motor Co., Ltd. | Information providing device, information providing method, and recording medium for presenting words extracted from different word groups |
US11115521B2 (en) | 2019-06-20 | 2021-09-07 | Verint Americas Inc. | Systems and methods for authentication and fraud detection |
US11538128B2 (en) | 2018-05-14 | 2022-12-27 | Verint Americas Inc. | User interface for fraud alert management |
US11868453B2 (en) | 2019-11-07 | 2024-01-09 | Verint Americas Inc. | Systems and methods for customer authentication based on audio-of-interest |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20200034020A (en) | 2018-09-12 | 2020-03-31 | 삼성전자주식회사 | Electronic apparatus and control method thereof |
US11805200B2 (en) * | 2022-01-25 | 2023-10-31 | International Business Machines Corporation | Detecting and resolving fraudulent calls |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7965822B2 (en) * | 2005-09-28 | 2011-06-21 | Siemens Enterprise Communications, Inc. | Telephony dialing using speech recognition |
US8000452B2 (en) * | 2004-07-26 | 2011-08-16 | General Motors Llc | Method and system for predictive interactive voice recognition |
US20120094647A1 (en) * | 2008-02-12 | 2012-04-19 | Microsoft Corporation | Utilizing mobile device functionality from remote computers |
-
2010
- 2010-03-18 US US12/726,932 patent/US8774372B2/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8000452B2 (en) * | 2004-07-26 | 2011-08-16 | General Motors Llc | Method and system for predictive interactive voice recognition |
US7965822B2 (en) * | 2005-09-28 | 2011-06-21 | Siemens Enterprise Communications, Inc. | Telephony dialing using speech recognition |
US20120094647A1 (en) * | 2008-02-12 | 2012-04-19 | Microsoft Corporation | Utilizing mobile device functionality from remote computers |
Cited By (51)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120263285A1 (en) * | 2005-04-21 | 2012-10-18 | Anthony Rajakumar | Systems, methods, and media for disambiguating call data to determine fraud |
US8793131B2 (en) | 2005-04-21 | 2014-07-29 | Verint Americas Inc. | Systems, methods, and media for determining fraud patterns and creating fraud behavioral models |
US8930261B2 (en) | 2005-04-21 | 2015-01-06 | Verint Americas Inc. | Method and system for generating a fraud risk score using telephony channel based audio and non-audio data |
US9113001B2 (en) * | 2005-04-21 | 2015-08-18 | Verint Americas Inc. | Systems, methods, and media for disambiguating call data to determine fraud |
US20100303211A1 (en) * | 2005-04-21 | 2010-12-02 | Victrio | Method and system for generating a fraud risk score using telephony channel based audio and non-audio data |
US9503571B2 (en) | 2005-04-21 | 2016-11-22 | Verint Americas Inc. | Systems, methods, and media for determining fraud patterns and creating fraud behavioral models |
US9571652B1 (en) | 2005-04-21 | 2017-02-14 | Verint Americas Inc. | Enhanced diarization systems, media and methods of use |
WO2012160193A1 (en) * | 2011-05-26 | 2012-11-29 | Jajah Ltd. | Voice conversation analysis utilising keywords |
ES2408906R1 (en) * | 2011-05-26 | 2013-08-06 | Telefonica Sa | SYSTEM AND METHOD FOR ANALYZING THE CONTENT OF A VOICE CONVERSATION |
US20140362738A1 (en) * | 2011-05-26 | 2014-12-11 | Telefonica Sa | Voice conversation analysis utilising keywords |
US9875739B2 (en) | 2012-09-07 | 2018-01-23 | Verint Systems Ltd. | Speaker separation in diarization |
US10950241B2 (en) | 2012-11-21 | 2021-03-16 | Verint Systems Ltd. | Diarization using linguistic labeling with segmented and clustered diarized textual transcripts |
US10438592B2 (en) | 2012-11-21 | 2019-10-08 | Verint Systems Ltd. | Diarization using speech segment labeling |
US11776547B2 (en) | 2012-11-21 | 2023-10-03 | Verint Systems Inc. | System and method of video capture and search optimization for creating an acoustic voiceprint |
US10950242B2 (en) | 2012-11-21 | 2021-03-16 | Verint Systems Ltd. | System and method of diarization and labeling of audio data |
US10720164B2 (en) | 2012-11-21 | 2020-07-21 | Verint Systems Ltd. | System and method of diarization and labeling of audio data |
US10692500B2 (en) | 2012-11-21 | 2020-06-23 | Verint Systems Ltd. | Diarization using linguistic labeling to create and apply a linguistic model |
US10134401B2 (en) | 2012-11-21 | 2018-11-20 | Verint Systems Ltd. | Diarization using linguistic labeling |
US10134400B2 (en) | 2012-11-21 | 2018-11-20 | Verint Systems Ltd. | Diarization using acoustic labeling |
US10692501B2 (en) | 2012-11-21 | 2020-06-23 | Verint Systems Ltd. | Diarization using acoustic labeling to create an acoustic voiceprint |
US11380333B2 (en) | 2012-11-21 | 2022-07-05 | Verint Systems Inc. | System and method of diarization and labeling of audio data |
US11367450B2 (en) | 2012-11-21 | 2022-06-21 | Verint Systems Inc. | System and method of diarization and labeling of audio data |
US10902856B2 (en) | 2012-11-21 | 2021-01-26 | Verint Systems Ltd. | System and method of diarization and labeling of audio data |
US10446156B2 (en) | 2012-11-21 | 2019-10-15 | Verint Systems Ltd. | Diarization using textual and audio speaker labeling |
US10522152B2 (en) | 2012-11-21 | 2019-12-31 | Verint Systems Ltd. | Diarization using linguistic labeling |
US10522153B2 (en) | 2012-11-21 | 2019-12-31 | Verint Systems Ltd. | Diarization using linguistic labeling |
US10650826B2 (en) | 2012-11-21 | 2020-05-12 | Verint Systems Ltd. | Diarization using acoustic labeling |
US11322154B2 (en) | 2012-11-21 | 2022-05-03 | Verint Systems Inc. | Diarization using linguistic labeling |
US11227603B2 (en) | 2012-11-21 | 2022-01-18 | Verint Systems Ltd. | System and method of video capture and search optimization for creating an acoustic voiceprint |
US10109280B2 (en) | 2013-07-17 | 2018-10-23 | Verint Systems Ltd. | Blind diarization of recorded calls with arbitrary number of speakers |
US9881617B2 (en) | 2013-07-17 | 2018-01-30 | Verint Systems Ltd. | Blind diarization of recorded calls with arbitrary number of speakers |
US9460722B2 (en) | 2013-07-17 | 2016-10-04 | Verint Systems Ltd. | Blind diarization of recorded calls with arbitrary number of speakers |
US10665253B2 (en) | 2013-08-01 | 2020-05-26 | Verint Systems Ltd. | Voice activity detection using a soft decision mechanism |
US11670325B2 (en) | 2013-08-01 | 2023-06-06 | Verint Systems Ltd. | Voice activity detection using a soft decision mechanism |
US9984706B2 (en) | 2013-08-01 | 2018-05-29 | Verint Systems Ltd. | Voice activity detection using a soft decision mechanism |
US10366693B2 (en) | 2015-01-26 | 2019-07-30 | Verint Systems Ltd. | Acoustic signature building for a speaker from multiple sessions |
US10726848B2 (en) | 2015-01-26 | 2020-07-28 | Verint Systems Ltd. | Word-level blind diarization of recorded calls with arbitrary number of speakers |
US9875742B2 (en) | 2015-01-26 | 2018-01-23 | Verint Systems Ltd. | Word-level blind diarization of recorded calls with arbitrary number of speakers |
US11636860B2 (en) | 2015-01-26 | 2023-04-25 | Verint Systems Ltd. | Word-level blind diarization of recorded calls with arbitrary number of speakers |
US9875743B2 (en) | 2015-01-26 | 2018-01-23 | Verint Systems Ltd. | Acoustic signature building for a speaker from multiple sessions |
US10657971B1 (en) | 2017-12-15 | 2020-05-19 | NortonLifeLock Inc. | Systems and methods for detecting suspicious voice calls |
US10971148B2 (en) * | 2018-03-30 | 2021-04-06 | Honda Motor Co., Ltd. | Information providing device, information providing method, and recording medium for presenting words extracted from different word groups |
US11538128B2 (en) | 2018-05-14 | 2022-12-27 | Verint Americas Inc. | User interface for fraud alert management |
CN108900342A (en) * | 2018-07-03 | 2018-11-27 | 阳光保险集团股份有限公司 | Telephone service processing method, telephone service processing unit and server |
CN109147801A (en) * | 2018-09-30 | 2019-01-04 | 深圳市元征科技股份有限公司 | voice interactive method, system, terminal and storage medium |
US11240372B2 (en) | 2018-10-25 | 2022-02-01 | Verint Americas Inc. | System architecture for fraud detection |
US10887452B2 (en) | 2018-10-25 | 2021-01-05 | Verint Americas Inc. | System architecture for fraud detection |
US20200162509A1 (en) * | 2018-11-19 | 2020-05-21 | OLX Global B.V. | Voice notes fraud detection |
US11652917B2 (en) | 2019-06-20 | 2023-05-16 | Verint Americas Inc. | Systems and methods for authentication and fraud detection |
US11115521B2 (en) | 2019-06-20 | 2021-09-07 | Verint Americas Inc. | Systems and methods for authentication and fraud detection |
US11868453B2 (en) | 2019-11-07 | 2024-01-09 | Verint Americas Inc. | Systems and methods for customer authentication based on audio-of-interest |
Also Published As
Publication number | Publication date |
---|---|
US8774372B2 (en) | 2014-07-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8774372B2 (en) | Telephone call inbox | |
US9858929B2 (en) | Computer-implemented system and method for transcription error reduction | |
US20210027247A1 (en) | Device, system and method for summarizing agreements | |
US10104233B2 (en) | Coaching portal and methods based on behavioral assessment data | |
US8990090B1 (en) | Script compliance using speech recognition | |
US20070133437A1 (en) | System and methods for enabling applications of who-is-speaking (WIS) signals | |
US10636047B2 (en) | System using automatically triggered analytics for feedback data | |
US9477971B2 (en) | Providing contextual information for spoken information | |
US9571652B1 (en) | Enhanced diarization systems, media and methods of use | |
US9930175B2 (en) | Systems and methods for lead routing | |
US8553871B2 (en) | Call monitoring system | |
US20170004178A1 (en) | Reference validity checker | |
US20140362738A1 (en) | Voice conversation analysis utilising keywords | |
US20050204378A1 (en) | System and method for video content analysis-based detection, surveillance and alarm management | |
JP6548974B2 (en) | Sales support information provision system and sales support information provision method | |
US11418647B1 (en) | Presenting multiple customer contact channels in a browseable interface | |
Qasim et al. | The Role of Customer Support Service in Relationship Strengthening: A Case of Swedish Broadband Internet Service Providers | |
US20080043953A1 (en) | System and method for collecting and processing caller information | |
CN113395395A (en) | AI-based intelligent voice return visit method | |
Murray | Voicemail greetings: a linguistic study | |
JP2016189510A (en) | Audio information storage device, audio information storage method and program |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: YEXT, INC., NEW YORK Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:METZ, BRENT D.;LERMAN, HOWARD;MACISAAC, SEAN;AND OTHERS;SIGNING DATES FROM 20100527 TO 20100601;REEL/FRAME:024486/0740 |
|
AS | Assignment |
Owner name: FELIX CALLS, LLC, NEW YORK Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:YEXT, INC.;REEL/FRAME:028781/0884 Effective date: 20120810 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YR, SMALL ENTITY (ORIGINAL EVENT CODE: M2551) Year of fee payment: 4 |
|
AS | Assignment |
Owner name: FELIX CALLS, LLC, NEW YORK Free format text: CHANGE OF NAME;ASSIGNOR:YEXT, INC.;REEL/FRAME:046040/0098 Effective date: 20120331 |
|
AS | Assignment |
Owner name: JPMORGAN CHASE BANK, N.A., NEW YORK Free format text: SECURITY AGREEMENT;ASSIGNOR:FELIX CALLS, LLC;REEL/FRAME:046432/0655 Effective date: 20180626 |
|
AS | Assignment |
Owner name: FELIX CALLS, LLC, NEW YORK Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:JPMORGAN CHASE BANK, N.A.;REEL/FRAME:047878/0531 Effective date: 20181231 |
|
AS | Assignment |
Owner name: COMERICA BANK, MICHIGAN Free format text: SECURITY INTEREST;ASSIGNOR:FELIX CALLS, LLC;REEL/FRAME:047890/0710 Effective date: 20181231 |
|
AS | Assignment |
Owner name: FELIX CALLS, LLC, PENNSYLVANIA Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:COMERICA BANK;REEL/FRAME:051128/0674 Effective date: 20191125 |
|
FEPP | Fee payment procedure |
Free format text: MAINTENANCE FEE REMINDER MAILED (ORIGINAL EVENT CODE: REM.); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY |
|
LAPS | Lapse for failure to pay maintenance fees |
Free format text: PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES (ORIGINAL EVENT CODE: EXP.); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY |
|
STCH | Information on status: patent discontinuation |
Free format text: PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362 |
|
FP | Lapsed due to failure to pay maintenance fee |
Effective date: 20220708 |